Signal generator not recognizing external timebase.

[PoDuck]

I have a Fluke 6060A/AN signal generator, and I'm trying to figure out what I need to do to get my timebase to be recognized as an external reference by it directly.  I have a GPSDO that produces a 10Mhz sine wave, and my frequency counters all seem to recognize it as an external time reference.  When I plug it into my signal generator though, the signal generator produces an uncal error, as though it isn't seeing a signal.  I would assume something is wrong with the input, but when I run a signal from the output of my frequency counter, which puts out what seems to be a terrible looking square wave, the signal generator recognizes it, and works fine.

When I plug the output of the frequency counter to my scope, it appears that the signal is all above zero, while the GPSDO puts out signal both above and below zero.  I'm not sure if that makes a difference or not, but I was wondering if anyone has any ideas about what might be the problem.

[4thDoctorWhoFan]

The obvious question is are you flipping the switch to EXT.?
Secondly, the manual states the EXT input needs to be a TTL timebase signal.  Are you using a TTL signal?

[PoDuck]

Well, it's not technically a TTL signal.  It's a regular sine wave, where the maximum peak is +21V, and the minimum peak is -21V, so it does cross the 0.8 and 2.4 volt thresholds.  Will this not work?

[PoDuck]

By the way, to answer your question, yes, I am flipping the switch to ext.  If I didn't, it would be using the internal timebase, and would not be giving me an uncal.

[PoDuck]

Sorry, lol, just looked, and I thought that signal was a little strong.  That was mV, not volts.

[DaJMasta]

Sorry, lol, just looked, and I thought that signal was a little strong.  That was mV, not volts.

So the signal is 42mV peak to peak....?  If so, that's tiny, and it's no surprise it can't see it.

A TTL signal is going to need at least a few volts to be readily accepted... it may also need enough DC bias to go from 0V to the peak without going negative, though depending on how they set it up inside, this may not be necessary.  If you've got the right amplitude and any bias that's required, a sine is probably fine.

[PoDuck]

Yeah, I hooked up a small amplifier to it, and as soon as I get the signal up to 3.3V or higher it works.  Now I'm trying to figure out options to boost the signal strength without hanging an amplifier off the thing.  The signal is coming out of a Nortel NTBW50AA.

[PoDuck]

I was going through a splitter before, but when I go directly from the output to my scope, it shows +1.92V to -1.92V.  I'm thinking I might be able to increase the power from the power supply and get a stronger signal.  I think it is supposed to be able to work from 24V to 48V, IIRC, but I'll have to double check that.

[DaJMasta]

It could be that your output signal is not driven for 50 ohm lines but for high impedance lines, which especially if you're splitting it (or if it has a couple physical outputs on the same output amp), could mean very low effective output power.

There are distribution amps that would repeat the signal (or amplify) over several outputs, but you could also give it some bias, shove it into a TTL inverter, and just get a squarewave output.  Lots of power in the harmonics, so it's not a pure tone like a sine wave, but for a reference for a PLL inside an external instrument, it works just fine, and it's very easy to build up and get a fair amount of output power.  Otherwise, in terms of designing something yourself, a dedicated output driver amp is a good choice, or an opamp with enough slew rate/gain bandwidth product to output the signal at the level you want.

[PoDuck]

I was considering just using one of these, which I have several of, and just rigging it up inside.  It's either that, or figure out how to offset the voltage to be all positive, which would give me +3.8V to 0V also.  The latter is probably cheaper, but more time consuming, and $10 is not terribly expensive.  Not sure which would be the better solution though.

[DaJMasta]

Something like that would work, but it would inject some noise into the signal - generally you want narrower band amps and a bandpass filter to keep from amplifying unwanted stray signals or harmonics.  Since it's only 10MHz, a simple transistor amp on some protoboard would do the trick too, could be an interesting little project.

As for biasing, it's actually really easy: put in a DC blocking capacitor on the input (to be sure), then use a resistor divider to set the bias voltage.  To keep the 10MHz out of your power lines, you could add an LC filter on the rail right before the top side of the divider.

[PoDuck]

Now I have to think about how much easier it would be to just always turn on my frequency counter when I am using my signal generator, and putting the GPSDO signal into the frequency counter, and taking the output of the frequency counter to the signal generator.  That is the lazy way to do it.  I just have to figure out how lazy I am.  lol

Anyway, thanks a lot for the help.